A good week-end working on the branh... I was able to get through a lot of issues, and have gone down to 13 failures and 2 errors, out of 573 tests (ie, I have 97% of the tests passing green ;-)
The remaining errors are mostly specific cases. For instance I found that we were allowing searches using a filter like '(street >= 3)' which is non-sensical, as 'street' does not have an ORDERDING MatchingRule. Otherwise, I had to fix a lot of small things in the API. >From the performance POV, a lookup done on a single entry is around 40% faster than what it was (in core-integ, I haven't yet checked the performance when going through the network). This is quite impressive... Sadly, the week-end is over, and I'll have to go back to work tomorrow. I hope to be able to fix the remaining errors in the next coming evenings.
